Objective: To update the existing CHIP (CT in Head Injury Patients) decision rule for detection of (in-tra)cranial findings in adult patients following minor head injury (MHI).Methods: The study is... Show moreObjective: To update the existing CHIP (CT in Head Injury Patients) decision rule for detection of (in-tra)cranial findings in adult patients following minor head injury (MHI).Methods: The study is a prospective multicenter cohort study in the Netherlands. Consecutive MHI pa-tients of 16 years and older were included. Primary outcome was any (intra)cranial traumatic finding on computed tomography (CT). Secondary outcomes were any potential neurosurgical lesion and neuro-surgical intervention. The CHIP model was validated and subsequently updated and revised. Diagnostic performance was assessed by calculating the c-statistic. Results: Among 4557 included patients 3742 received a CT (82%). In 383 patients (8.4%) a traumatic find-ing was present on CT. A potential neurosurgical lesion was found in 73 patients (1.6%) with 26 (0.6%) patients that actually had neurosurgery or died as a result of traumatic brain injury. The original CHIP underestimated the risk of traumatic (intra)cranial findings in low-predicted-risk groups, while in high -predicted-risk groups the risk was overestimated. The c-statistic of the original CHIP model was 0.72 (95% CI 0.69-0.74) and it would have missed two potential neurosurgical lesions and one patient that underwent neurosurgery. The updated model performed similar to the original model regarding trau-matic (intra)cranial findings (c-statistic 0.77 95% CI 0.74-0.79, after crossvalidation c-statistic 0.73). The updated CHIP had the same CT rate as the original CHIP (75%) and a similar sensitivity (92 versus 93%) and specificity (both 27%) for any traumatic (intra)cranial finding. However, the updated CHIP would not have missed any (potential) neurosurgical lesions and had a higher sensitivity for (potential) neurosurgi-cal lesions or death as a result of traumatic brain injury (100% versus 96%).Conclusions: Use of the updated CHIP decision rule is a good alternative to current decision rules for patients with MHI. In contrast to the original CHIP the update identified all patients with (potential) neurosurgical lesions without increasing CT rate.(c) 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ) Show less
Background: While the Glasgow coma scale (GCS) is one of the strongest outcome predictors, the current classification of traumatic brain injury (TBI) as'mild" 'moderate'or'severe' based on this... Show moreBackground: While the Glasgow coma scale (GCS) is one of the strongest outcome predictors, the current classification of traumatic brain injury (TBI) as'mild" 'moderate'or'severe' based on this fails to capture enormous heterogeneity in pathophysiology and treatment response. We hypothesized that data-driven characterization of TBl could identify distinct endotypes and give mechanistic insights. Methods: We developed an unsupervised statistical clustering model based on a mixture of probabilistic graphs for presentation (<24 h) demographic, clinical, physiological, laboratory and imaging data to identify subgroups of TBl patients admitted to the intensive care unit in the CENTER-TBI dataset (N= 1,728). A cluster similarity index was used for robust determination of optimal cluster number. Mutual information was used to quantify feature importance and for cluster interpretation. Results: Six stable endotypes were identified with distinct GCS and composite systemic metabolic stress profiles, distinguished by GCS, blood lactate, oxygen saturation, serum creatinine, glucose, base excess, pH, arterial partial pressure of carbon dioxide, and body temperature. Notably, a cluster with 'moderate'TBI (by traditional classification) and deranged metabolic profile, had a worse outcome than a cluster with 'severe'GCS and a normal metabolic profile. Addition of cluster labels significantly improved the prognostic precision of the IMPACT (International Mission for Prognosis and Analysis of Clinical trials in TBI) extended model, for prediction of both unfavourable outcome and mortality (both p <0.001). Conclusions: Six stable and clinically distinct TBI endotypes were identified by probabilistic unsupervised clustering. In addition to presenting neurology, a profile of biochemical derangement was found to be an important distinguishing feature that was both biologically plausible and associated with outcome. Our work motivates refining current TBI classifications with factors describing metabolic stress. Such data-driven clusters suggest TBI endotypes that merit investigation to identify bespoke treatment strategies to improve care. Show less
Background Evaluating patients' experiences is essential when incorporating the patients' perspective in improving healthcare. Experiences are mainly collected using closed-ended questions,... Show moreBackground Evaluating patients' experiences is essential when incorporating the patients' perspective in improving healthcare. Experiences are mainly collected using closed-ended questions, although the value of open-ended questions is widely recognized. Natural language processing (NLP) can automate the analysis of open-ended questions for an efficient approach to patient-centeredness. Methods We developed the Artificial Intelligence Patient-Reported Experience Measures (AI-PREM) tool, consisting of a new, open-ended questionnaire, an NLP pipeline to analyze the answers using sentiment analysis and topic modeling, and a visualization to guide physicians through the results. The questionnaire and NLP pipeline were iteratively developed and validated in a clinical context. Results The final AI-PREM consisted of five open-ended questions about the provided information, personal approach, collaboration between healthcare professionals, organization of care, and other experiences. The AI-PREM was sent to 867 vestibular schwannoma patients, 534 of which responded. The sentiment analysis model attained an F1 score of 0.97 for positive texts and 0.63 for negative texts. There was a 90% overlap between automatically and manually extracted topics. The visualization was hierarchically structured into three stages: the sentiment per question, the topics per sentiment and question, and the original patient responses per topic. Conclusions The AI-PREM tool is a comprehensive method that combines a validated, open-ended questionnaire with a well-performing NLP pipeline and visualization. Thematically organizing and quantifying patient feedback reduces the time invested by healthcare professionals to evaluate and prioritize patient experiences without being confined to the limited answer options of closed-ended questions. Show less
When a patient is admitted to the intensive care unit (ICU) after a traumatic brain injury (TBI), an early prognosis is essential for baseline risk adjustment and shared decision making. TBI... Show moreWhen a patient is admitted to the intensive care unit (ICU) after a traumatic brain injury (TBI), an early prognosis is essential for baseline risk adjustment and shared decision making. TBI outcomes are commonly categorised by the Glasgow Outcome Scale-Extended (GOSE) into eight, ordered levels of functional recovery at 6 months after injury. Existing ICU prognostic models predict binary outcomes at a certain threshold of GOSE (e.g., prediction of survival [GOSE > 1]). We aimed to develop ordinal prediction models that concurrently predict probabilities of each GOSE score. From a prospective cohort (n = 1,550, 65 centres) in the ICU stratum of the Collaborative European NeuroTrauma Effectiveness Research in TBI (CENTER-TBI) patient dataset, we extracted all clinical information within 24 hours of ICU admission (1,151 predictors) and 6-month GOSE scores. We analysed the effect of two design elements on ordinal model performance: (1) the baseline predictor set, ranging from a concise set of ten validated predictors to a token-embedded representation of all possible predictors, and (2) the modelling strategy, from ordinal logistic regression to multinomial deep learning. With repeated k-fold cross-validation, we found that expanding the baseline predictor set significantly improved ordinal prediction performance while increasing analytical complexity did not. Half of these gains could be achieved with the addition of eight high-impact predictors to the concise set. At best, ordinal models achieved 0.76 (95% CI: 0.74-0.77) ordinal discrimination ability (ordinal c-index) and 57% (95% CI: 54%- 60%) explanation of ordinal variation in 6-month GOSE (Somers' D-xy). Model performance and the effect of expanding the predictor set decreased at higher GOSE thresholds, indicating the difficulty of predicting better functional outcomes shortly after ICU admission. Our results motivate the search for informative predictors that improve confidence in prognosis of higher GOSE and the development of ordinal dynamic prediction models. Show less
Ullah, W.; Steyerberg, E.W.; Tchantchaleishvili, V. 2022
Retrospective healthcare databases are emerging sources for clinical research. However, there has been no standardized checklist to ensure the accurate acquisition and reporting of this data. We... Show moreRetrospective healthcare databases are emerging sources for clinical research. However, there has been no standardized checklist to ensure the accurate acquisition and reporting of this data. We consulted experts, statisticians and searched for digital databases to develop a comprehensive checklist with a focus on the issues specific to studies performed on retrospective databases. The ChARDS (Checklist for Administrative and Research Databases related Studies) was developed that consists of 8 sections, 19 sub-sections, and 57 questions. The major areas covered by the ChARDS include providing information on the data sources and guiding writing by simulating the headings of a manuscript. The ChARDS is designed to question authors on the need for the study, the relevance of the topic in light of prior literature, research design, selection of the sample, eligibility of participants, standardization of outcomes, appropriateness of statistical models, interpretation of data, resource valuation, reliability and reproducibility of results, and validity and generalization of key findings to the general population. The ChARDS intends to provide authors with a roadmap on the structured reporting of data and enable decision-makers to evaluate its suitability for publication. Show less
Background Despite being well established, acute surgery in traumatic acute subdural haematoma is based on low-grade evidence. We aimed to compare the effectiveness of a strategy preferring acute... Show moreBackground Despite being well established, acute surgery in traumatic acute subdural haematoma is based on low-grade evidence. We aimed to compare the effectiveness of a strategy preferring acute surgical evacuation with one preferring initial conservative treatment in acute subdural haematoma.Methods We did a prospective, observational, comparative effectiveness study using data from participants enrolled in the Collaborative European Neurotrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) cohort. We included patients with no pre-existing severe neurological disorders who presented with acute subdural haematoma within 24 h of traumatic brain injury. Using an instrumental variable analysis, we compared outcomes between centres according to treatment preference for acute subdural haematoma (acute surgical evacuation or initial conservative treatment), measured by the case-mix-adjusted percentage of acute surgery per centre. The primary endpoint was functional outcome at 6 months as rated with the Glasgow Outcome Scale Extended, which was estimated with ordinal regression as a common odds ratio (OR) and adjusted for prespecified confounders. Variation in centre preference was quantified with the median OR (MOR). CENTER-TBI is registered with ClinicalTrials.gov , number NCT02210221, and the Resource Identification Portal (Research Resource Identifier SCR_015582).Findings Between Dec 19, 2014 and Dec 17, 2017, 4559 patients with traumatic brain injury were enrolled in CENTER-TBI, of whom 1407 (31%) presented with acute subdural haematoma and were included in our study. Acute surgical evacuation was done in 336 (24%) patients, by craniotomy in 245 (73%) of those patients and by decompressive craniectomy in 91 (27%). Delayed decompressive craniectomy or craniotomy after initial conservative treatment (n=982) occurred in 107 (11%) patients. The percentage of patients who underwent acute surgery ranged from 5.6% to 51.5% (IQR 12.3-35.9) between centres, with a two-times higher probability of receiving acute surgery for an identical patient in one centre versus another centre at random (adjusted MOR for acute surgery 1.8; p<0.0001]). Centre preference for acute surgery over initial conservative treatment was not associated with improvements in functional outcome (common OR per 23.6% [IQR increase] more acute surgery in a centre 0.92, 95% CI 0.77-1.09).Interpretation Our findings show that treatment for patients with acute subdural haematoma with similar characteristics differed depending on the treating centre, because of variation in the preferred approach. A treatment strategy preferring an aggressive approach of acute surgical evacuation over initial conservative treatment was not associated with better functional outcome. Therefore, in a patient with acute subdural haematoma for whom a neurosurgeon sees no clear superiority for acute surgery over conservative treatment, initial conservative treatment might be considered. Copyright (C) 2022 Published by Elsevier Ltd. All rights reserved. Show less
Simple Summary Barrett's esophagus (BE) is the only known precursor lesion of esophageal adenocarcinoma (EAC). Endoscopic surveillance plays an important role in the timely detection of neoplastic... Show moreSimple Summary Barrett's esophagus (BE) is the only known precursor lesion of esophageal adenocarcinoma (EAC). Endoscopic surveillance plays an important role in the timely detection of neoplastic progression. However, the cost-effectiveness of current surveillance strategies is debatable. Previous studies have shown that male Barrett's patients have lower neoplastic progression risk than females. However, these studies do not provide a more practical translation of these sex disparities into different surveillance intervals. The current multicenter prospective cohort study aimed to evaluate sex differences in 868 BE patients; not only with respect to neoplastic progression risk, but also concerning the difference in time to detection of high-grade dysplasia (HGD)/EAC: time to neoplastic progression was estimated to be almost twice as low in males than in females. In contrast, the stage of neoplasia appeared to be higher in females. Our results can guide future discussions for sex-specific guidelines, supporting the implementation of neoplastic risk stratification per individual patient in BE surveillance. Recommendations in Barrett's esophagus (BE) guidelines are mainly based on male patients. We aimed to evaluate sex differences in BE patients in (1) probability of and (2) time to neoplastic progression, and (3) differences in the stage distribution of neoplasia. We conducted a multicenter prospective cohort study including 868 BE patients. Cox regression modeling and accelerated failure time modeling were used to estimate the sex differences. Neoplastic progression was defined as high-grade dysplasia (HGD) and/or esophageal adenocarcinoma (EAC). Among the 639 (74%) males and 229 females that were included (median follow-up 7.1 years), 61 (7.0%) developed HGD/EAC. Neoplastic progression risk was estimated to be twice as high among males (HR 2.26, 95% CI 1.11-4.62) than females. The risk of HGD was found to be higher in males (HR 3.76, 95% CI 1.33-10.6). Time to HGD/EAC (AR 0.52, 95% CI 0.29-0.95) and HGD (AR 0.40, 95% CI 0.19-0.86) was shorter in males. Females had proportionally more EAC than HGD and tended to have higher stages of neoplasia at diagnosis. In conclusion, both the risk of and time to neoplastic progression were higher in males. However, females were proportionally more often diagnosed with (advanced) EAC. We should strive for improved neoplastic risk stratification per individual BE patient, incorporating sex disparities into new prediction models. Show less
Dregmans, E.; Kaal, A.G.; Meziyerh, S.; Kolfschoten, N.E.; Aken, M.O. van; Schippers, E.F.; ... ; Nieuwkoop, C. van 2022
IMPORTANCE Misdiagnosis of infection is among the most commonly made diagnostic errors and is associated with increased morbidity and mortality. Little is known about how often misdiagnosed site of... Show moreIMPORTANCE Misdiagnosis of infection is among the most commonly made diagnostic errors and is associated with increased morbidity and mortality. Little is known about how often misdiagnosed site of infection occurs and its association with clinical outcomes.OBJECTIVES To evaluate the discrepancy between admission and discharge site of infection diagnoses among patients with suspected bacteremia, to explore factors associated with discrepant diagnoses, and to evaluate the association with clinical outcomes.DESIGN, SETTING, AND PARTICIPANTS This cohort study used electronic records of 1477 adult patients who were admitted to the hospital for suspected bacteremia from April 1, 2019, to May 31, 2020, and who had blood cultures taken at the emergency department at Haga Teaching Hospital, The Hague, the Netherlands. Suspected infection sites were classified into 8 categories at admission and discharge. Misdiagnosed site was defined as a discrepancy between the suspected site of infection at admission and at discharge.MAIN OUTCOMES AND MEASURES Clinical outcomes were 30-day mortality, intensive care unit admission, length of hospital stay, and antibiotic use, analyzed with logistic and linear regression. Risk factors for misdiagnosed site were determined using regression analysis.RESULTS A total of 1477 patients (820 [55.5%] male; median [IQR] age, 68 [56-78] years) were analyzed. The rate of misdiagnosed site of infection was 11.6% (171 of 1477); 3.1% of all patients (46 of 1477) ultimately had no infection. No association was found between misdiagnosis and 30-day mortality (adjusted odds ratio [aOR], 0.8; 95% CI, 03-1.9; P = .60), intensive care unit admission (aOR, 1.3; 95% CI, 0.6-3.0; P = .54), and hospital length of stay (adjusted increase of stay, 15.5%; 95% CI, -3.1% to 37.7%; P = .11). Misdiagnosed site was associated with receiving broad-spectrum antibiotics (aOR, 4.0; 95% CI, 1.8-8.8; P < .001). Older age, dementia, a positive urine sediment test result without urinary symptoms, and suspicion of an intravascular, central nervous system, or bone and joint infection were risk factors for misdiagnosed site of infection.CONCLUSIONS AND RELEVANCE In this cohort study, misdiagnosed site of infection occurred in 1 of 9 patients and was not associated with worse short-term clinical outcomes. Clinicians should be aware of risk factors associated with misdiagnosed site of infection and potential inappropriate antibiotic use. Show less
Machine learning is increasingly being used to predict clinical outcomes. Most comparisons of different methods have been based on empirical analyses in specific datasets. We used Monte Carlo... Show moreMachine learning is increasingly being used to predict clinical outcomes. Most comparisons of different methods have been based on empirical analyses in specific datasets. We used Monte Carlo simulations to determine when machine learning methods perform better than statistical learning methods in a specific setting. We evaluated six learning methods: stochastic gradient boosting machines using trees as the base learners, random forests, artificial neural networks, the lasso, ridge regression, and linear regression estimated using ordinary least squares (OLS). Our simulations were informed by empirical analyses in patients with acute myocardial infarction (AMI) and congestive heart failure (CHF) and used six data-generating processes, each based on one of the six learning methods, to simulate continuous outcomes in the derivation and validation samples. The outcome was systolic blood pressure at hospital discharge, a continuous outcome. We applied the six learning methods in each of the simulated derivation samples and evaluated performance in the simulated validation samples. The primary observation was that neural networks tended to result in estimates with worse predictive accuracy than the other five methods in both disease samples and across all six data-generating processes. Boosted trees and OLS regression tended to perform well across a range of scenarios. Show less
Objective: To compare outcomes between patients with primary external ventricular device (EVD)-driven treatment of intracranial hypertension and those with primary intraparenchymal monitor (IP)... Show moreObjective: To compare outcomes between patients with primary external ventricular device (EVD)-driven treatment of intracranial hypertension and those with primary intraparenchymal monitor (IP)-driven treatment.Methods: The CENTER-TBI study is a prospective, multicenter, longitudinal observational cohort study that enrolled patients of all TBI severities from 62 participating centers (mainly level I trauma centers) across Europe between 2015 and 2017. Functional outcome was assessed at 6 months and a year. We used multivariable adjusted instrumental variable (IV) analysis with "center" as instrument and logistic regression with covariate adjustment to determine the effect estimate of EVD on 6-month functional outcome. Results: A total of 878 patients of all TBI severities with an indication for intracranial pressure (ICP) monitoring were included in the present study, of whom 739 (84%) patients had an IP monitor and 139 (16%) an EVD. Patients included were predominantly male (74% in the IP monitor and 76% in the EVD group), with a median age of 46 years in the IP group and 48 in the EVD group. Six-month GOS-E was similar between IP and EVD patients (adjusted odds ratio (aOR) and 95% confidence interval [CI] OR 0.74 and 95% CI [0.36-1.52], adjusted IV analysis). The length of intensive care unit stay was greater in the EVD group than in the IP group (adjusted rate ratio [95% CI] 1.70 [1.34-2.12], IV analysis). One hundred eighty-seven of the 739 patients in the IP group (25%) required an EVD due to refractory ICPs. Conclusion: We found no major differences in outcomes of patients with TBI when comparing EVD-guided and IP monitor-guided ICP management. In our cohort, a quarter of patients that initially received an IP monitor required an EVD later for ICP control. The prevalence of complications was higher in the EVD group. Show less
Traumatic brain injury (TBI) remains one of the most fatal and debilitating conditions in the world. Current clinical management in severe TBI patients is mainly concerned with reducing secondary... Show moreTraumatic brain injury (TBI) remains one of the most fatal and debilitating conditions in the world. Current clinical management in severe TBI patients is mainly concerned with reducing secondary insults and optimizing the balance between substrate delivery and consumption. Over the past decades, multimodality monitoring has become more widely available, and clinical management protocols have been published that recommend potential interventions to correct pathophysiological derangements. Even while evidence from randomized clinical trials is still lacking for many of the recommended interventions, these protocols and algorithms can be useful to define a clear standard of therapy where novel interventions can be added or be compared to. Over the past decade, more attention has been paid to holistic management, in which hemodynamic, respiratory, inflammatory or coagulation disturbances are detected and treated accordingly. Considerable variability with regards to the trajectories of recovery exists. Even while most of the recovery occurs in the first months after TBI, substantial changes may still occur in a later phase. Neuroprognostication is challenging in these patients, where a risk of self-fulfilling prophecies is a matter of concern. The present article provides a comprehensive and practical review of the current best practice in clinical management and long-term outcomes of moderate to severe TBI in adult patients admitted to the intensive care unit. Show less
Traumatic brain injury is associated with changes to the metabolome. Here the authors show that acute traumatic brain injury has distinctive serum metabolic patterns which may suggest protective... Show moreTraumatic brain injury is associated with changes to the metabolome. Here the authors show that acute traumatic brain injury has distinctive serum metabolic patterns which may suggest protective changes of systemic lipid metabolism aiming to maintain lipid homeostasis in the brain.Complex metabolic disruption is a crucial aspect of the pathophysiology of traumatic brain injury (TBI). Associations between this and systemic metabolism and their potential prognostic value are poorly understood. Here, we aimed to describe the serum metabolome (including lipidome) associated with acute TBI within 24 h post-injury, and its relationship to severity of injury and patient outcome. We performed a comprehensive metabolomics study in a cohort of 716 patients with TBI and non-TBI reference patients (orthopedic, internal medicine, and other neurological patients) from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury (CENTER-TBI) cohort. We identified panels of metabolites specifically associated with TBI severity and patient outcomes. Choline phospholipids (lysophosphatidylcholines, ether phosphatidylcholines and sphingomyelins) were inversely associated with TBI severity and were among the strongest predictors of TBI patient outcomes, which was further confirmed in a separate validation dataset of 558 patients. The observed metabolic patterns may reflect different pathophysiological mechanisms, including protective changes of systemic lipid metabolism aiming to maintain lipid homeostasis in the brain. Show less
Background: Following traumatic brain injury (TBI), the clinical focus is often on disability. However, patients' perceptions of well-being can be discordant with their disability level, referred... Show moreBackground: Following traumatic brain injury (TBI), the clinical focus is often on disability. However, patients' perceptions of well-being can be discordant with their disability level, referred to as the 'disability paradox'. We aimed to examine the relationship between disability and health-related quality of life (HRQoL) following TBI, while taking variation in personal, injury-related and environment factors into account. Methods: We used data from the Collaborative European NeuroTrauma Effectiveness Research in Traumatic Brain Injury study. Disability was assessed 6 months post-injury by the Glasgow Outcome Scale-Extended (GOSE). HRQoL was assessed by the SF-12v2 physical and mental component summary scores and the Quality of Life after Traumatic Brain Injury overall scale. We examined mean total and domain HRQoL scores by GOSE. We quantified variance in HRQoL explained by GOSE, personal, injury-related and environment factors with multivariable regression. Results: Six-month outcome assessments were completed in 2075 patients, of whom 78% had mild TBI (Glasgow Coma Scale 13-15). Patients with severe disability had higher HRQoL than expected on the basis of GOSE alone, particularly after mild TBI. Up to 50% of patients with severe disability reported HRQoL scores within the normative range. GOSE, personal, injury-related and environment factors explained a limited amount of variance in HRQoL (up to 29%). Conclusion: Contrary to the idea that discrepancies are unusual, many patients with poor functional outcomes reported well-being that was at or above the boundary considered satisfactory for the normative sample. These findings challenge the idea that satisfactory HRQoL in patients with disability should be described as 'paradoxical' and question common views of what constitutes 'unfavourable' outcome. Show less
Large and complex data sets are increasingly available for research in critical care. To analyze these data, researchers use techniques commonly referred to as statistical learning or machine... Show moreLarge and complex data sets are increasingly available for research in critical care. To analyze these data, researchers use techniques commonly referred to as statistical learning or machine learning (ML). The latter is known for large successes in the field of diagnostics, for example, by identification of radiological anomalies. In other research areas, such as clustering and prediction studies, there is more discussion regarding the benefit and efficiency of ML techniques compared with statistical learning. In this viewpoint, we aim to explain commonly used statistical learning and ML techniques and provide guidance for responsible use in the case of clustering and prediction questions in critical care. Clustering studies have been increasingly popular in critical care research, aiming to inform how patients can be characterized, classified, or treated differently. An important challenge for clustering studies is to ensure and assess generalizability. This limits the application of findings in these studies toward individual patients. In the case of predictive questions, there is much discussion as to what algorithm should be used to most accurately predict outcome. Aspects that determine usefulness of ML, compared with statistical techniques, include the volume of the data, the dimensionality of the preferred model, and the extent of missing data. There are areas in which modern ML methods may be preferred. However, efforts should be made to implement statistical frameworks (e.g., for dealing with missing data or measurement error, both omnipresent in clinical data) in ML methods. To conclude, there are important opportunities but also pitfalls to consider when performing clustering or predictive studies with ML techniques. We advocate careful valuation of new data-driven findings. More interaction is needed between the engineer mindset of experts in ML methods, the insight in bias of epidemiologists, and the probabilistic thinking of statisticians to extract as much information and knowledge from data as possible, while avoiding harm. Show less
BACKGROUND: Missing data is a typical problem in clinical studies, where the value of variables of interest is not measured or collected for some patients. This article aimed to review imputation... Show moreBACKGROUND: Missing data is a typical problem in clinical studies, where the value of variables of interest is not measured or collected for some patients. This article aimed to review imputation approaches for missing values and their application in neurosurgery.METHODS: We reviewed current practices on detecting missingness patterns and applications of multiple imputation approaches under different scenarios. Statistical considerations and importance of sensitivity analysis were explained. Various imputation methods were applied to a retrospective cohort.RESULTS: For illustration purposes, a retrospective cohort of 609 patients harboring both ruptured and unruptured intracranial aneurysms and undergoing microsurgical clip reconstruction at Erasmus MC University Medical Center, Rotterdam, The Netherlands, between 2000 and 2019 was used. modified Rankin Scale score at 6 months was the clinical outcome, and potential predictors were age, sex, size of aneurysm, hypertension, smoking, World Federation of Neurosurgical Societies grade, and aneurysm location. Associations were investigated using different imputation approaches, and the results were compared and discussed.CONCLUSIONS: Missing values should be treated carefully. Advantages and disadvantages of multiple imputation methods along with imputation in small and big data should be considered depending on the research question and specifics of the study. Show less
Objectives: The current surveillance strategy in Barrett's esophagus (BE) uses only histological findings of the last endoscopy to assess neoplastic progression risk. As predictor values vary... Show moreObjectives: The current surveillance strategy in Barrett's esophagus (BE) uses only histological findings of the last endoscopy to assess neoplastic progression risk. As predictor values vary across endoscopies, single measurements may not be an accurate reflection. Our aim was to explore the value of using longitudinal evolutions (i.e. successive measurements) of histological findings (low-grade dysplasia (LGD)) and immunohistochemical biomarkers (p53 and SOX2) by investigating the association with Barrett's progression. Methods: In this proof-of-principle study of a longitudinal dynamic risk estimation model with a multicenter cohort design, 631 BE patients from 15 Dutch hospitals who were under surveillance were included. Longitudinal dynamic values of LGD, p53, and SOX2 were included in a multivariate joint model to estimate the risk of high-grade dysplasia (HGD)/esophageal adenocarcinoma (EAC). Results: Longitudinal evolutions of aberrant expression of p53 (HR 1.26, p < 0.01) and SOX2 (HR 1.43, p < 0.01) were associated with an increased HGD/EAC risk. We also found weak evidence of an association with the longitudinal evolution of the presence of LGD (HR 1.02, p = 0.12). The performance of the model was good (AUC 0.80-0.88). Using this model, for each future BE patient the probability of aberrant expression of biomarkers based on multiple longitudinal observations can be estimated. This probability is translated in progression risk, expressed as HR. Conclusions: This study provides solid ground to further explore a paradigm shift from currently recommended fixed intervals towards personalized surveillance, in which tailored risk estimations and corresponding surveillance intervals can be updated at every FU endoscopy for individual BE patients. Show less
Ceyisakar, I.E.; Leeuwen, N. van; Steyerberg, E.W.; Lingsma, H.F. 2022
Background: Instrumental variable (IV) analysis holds the potential to estimate treatment effects from observational data. IV analysis potentially circumvents unmeasured confounding but makes a... Show moreBackground: Instrumental variable (IV) analysis holds the potential to estimate treatment effects from observational data. IV analysis potentially circumvents unmeasured confounding but makes a number of assumptions, such as that the IV shares no common cause with the outcome. When using treatment preference as an instrument, a common cause, such as a preference regarding related treatments, may exist. We aimed to explore the validity and precision of a variant of IV analysis where we additionally adjust for the provider: adjusted IV analysis. Methods: A treatment effect on an ordinal outcome was simulated (beta - 0.5 in logistic regression) for 15.000 patients, based on a large data set (the IMPACT data, n = 8799) using different scenarios including measured and unmeasured confounders, and a common cause of IV and outcome. We compared estimated treatment effects with patient-level adjustment for confounders, IV with treatment preference as the instrument, and adjusted IV, with hospital added as a fixed effect in the regression models. Results: The use of patient-level adjustment resulted in biased estimates for all the analyses that included unmeasured confounders, IV analysis was less confounded, but also less reliable. With correlation between treatment preference and hospital characteristics (a common cause) estimates were skewed for regular IV analysis, but not for adjusted IV analysis. Conclusion: When using IV analysis for comparing hospitals, some limitations of regular IV analysis can be overcome by adjusting for a common cause. Show less
Background: The aim of this study was to develop a prediction model for 10-year overall survival (OS) after resection of colorectal liver metastasis (CRLM) based on patient, tumour and treatment... Show moreBackground: The aim of this study was to develop a prediction model for 10-year overall survival (OS) after resection of colorectal liver metastasis (CRLM) based on patient, tumour and treatment characteristics.Methods: Consecutive patients after complete resection of CRLM were included from two centres (1992-2019). A prediction model providing 10-year OS probabilities was developed using Cox regression analysis, including KRAS, BRAF and histopathological growth patterns. Discrimination and calibration were assessed using cross-validation. A web-based calculator was built to predict individual 10-year OS probabilities.Results: A total of 4112 patients were included. The estimated 10-year OS was 30% (95% CI 29 -32). Fifteen patient, tumour and treatment characteristics were independent prognostic factors for 10-year OS; age, gender, location and nodal status of the primary tumour, disease-free interval, number and diameter of CRLM, preoperative CEA, resection margin, extrahepatic disease, KRAS and BRAF mutation status, histopathological growth patterns, perioperative systemic chemotherapy and hepatic arterial infusion pump chemotherapy. The discrimination at 10-years was 0.73 for both centres. A simplified risk score identified four risk groups with a 10-year OS of 57%, 38%, 24%, and 12%.Conclusions: Ten-year OS after resection of CRLM is best predicted with a model including 15 patient, tumour, and treatment characteristics. The web-based calculator can be used to inform patients. This model serves as a benchmark to determine the prognostic value of novel biomarkers. (C) 2022 The Author(s). Published by Elsevier Ltd. Show less
Rietbergen, T.; Marang-van de Mheen, P.J.; Graaf, J. de; Diercks, R.L.; Janssen, R.P.A.; Linden-van der Zwaag, H.M.J. van der; ... ; SMART Study Grp 2022
Purpose: To evaluate the effectiveness of a tailored intervention to reduce low value MRIs and arthroscopies among patients >= 50 years with degenerative knee disease in 13 Dutch orthopaedic... Show morePurpose: To evaluate the effectiveness of a tailored intervention to reduce low value MRIs and arthroscopies among patients >= 50 years with degenerative knee disease in 13 Dutch orthopaedic centers (intervention group) compared with all other Dutch orthopaedic centers (control group). Methods: All patients with degenerative knee disease >= 50 years admitted to Dutch orthopaedic centers from January 2016 to December 2018 were included. The tailored intervention included participation of clinical champions, education on the Dutch Choosing Wisely recommendation for MRI's and arthroscopies in degenerative knee disease, training of orthopaedic surgeons to manage patient expectations, performance feedback, and provision of a patient brochure. A difference-in-difference analysis was used to compare the time trend before (admitted January 2016-June 2017) and after introduction of the intervention (July 2017-December 2018) between intervention and control hospitals. Primary outcome was the monthly percentage of patients receiving a MRI or knee arthroscopy, weighted by type of hospital. Results: 136,446 patients were included, of whom 32,163 were treated in the intervention hospitals. The weighted percentage of patients receiving a MRI on average declined by 0.15% per month (beta = - 0.15, P < 0.001) and by 0.19% per month for arthroscopy (beta = - 0.19, P < 0.001). However, these changes over time did not differ between intervention and control hospitals, neither for MRI (beta = - 0.74, P = 0.228) nor arthroscopy (beta = 0.13, P = 0.688). Conclusions: The extent to which patients >= 50 years with degenerative knee disease received a MRI or arthroscopy declined significantly over time, but could not be attributed to the tailored intervention. This secular downward time trend may reflect anoverall focus of reducing low value care in The Netherlands. Show less